I build these systems and I will still tell you that many owners asking this question should not buy a CRM yet.
That is especially true of a coaching or consulting practice, a small firm, or a clinic still running on the owner's memory. Not because the software is bad. Because the question is usually a symptom of something else, and buying a tool to answer it tends to produce a tidy, expensive record of the problem you already had.
The actual test
Forget software for a moment. Here is the diagnostic.
You are unavailable for three weeks. Somebody you trust has to speak to a client on your behalf. Can they open one place and find enough not to embarrass you? What the person came for, what has happened, what was promised, where things stand?
If the answer is no, you have a memory problem, and the memory problem is real whether or not you buy anything. If the answer is yes, in a spreadsheet, then your spreadsheet is a CRM and it is working.
The word CRM describes a job, not a purchase.
What most businesses actually need first
In order, and only move to the next when the current one genuinely stops holding.
One list of every person, in one file. Not four: a booking system, a mailing list, a notebook and your inbox. One place with everyone in it, including the people who inquired and never became clients, who are usually the most valuable names you own and are almost never anywhere.
Context on each name. What they came for in their own words, what was happening in their life or business at the time, what they have already had from you. This is the part that decides whether the record is usable, and it has nothing to do with software. I have written more on what belongs in a client record.
A commitment log. What was promised, by whom, by when. Three columns. This one thing prevents more quiet client loss than any tool, because service businesses rarely lose people through bad work. They lose them through a promise made in a good conversation that then did not happen.
A stage per person. Beginning, middle, finishing, finished, dormant. One word, so you can answer "who is drifting" without reading forty files.
A spreadsheet does all four. If yours does them and you keep it current, you do not have a problem to solve.
When a CRM starts earning its cost
Four conditions. I would want at least two before spending money.
More than one person needs to see it. The moment an assistant, an associate or a colleague has to know what is happening, a spreadsheet starts producing version conflicts and quiet overwrites. In a business with a team this condition is usually already met.
Something has to happen automatically. A reminder, a sequence, a nudge at day ten. Once the follow-up depends on you remembering, it does not exist, and remembering is worst exactly when you are busiest.
The volume has passed what you can hold. With twelve active clients you genuinely remember. With sixty names across active, dormant and inquired, you do not, and you have probably already lost several without noticing.
You need to see the shape rather than the people. How many inquiries stalled at the same stage, what happened to everyone from last year's workshop. Spreadsheets answer this badly.
Notice that none of these is "everyone says I should have one."
Why buying early backfires
Two reasons, and both are common enough that I now warn people.
A CRM does not create the discipline of writing things down. It relocates it. An owner who does not currently record what a client was dealing with will not start because a field exists, and the empty field is more demoralizing than the empty notebook was.
And the setup cost is paid in exactly the resource that was already short. Migration, configuration and learning consume the evenings you would have spent on the thing that would actually have changed the business. I have watched people spend a month on this and end up with the same business, better labelled.
If you take one thing: get the four items above working somewhere ugly first. If you still want software after three months of that, you will know exactly what you need it to do, which makes choosing trivial. If someone else is going to build it for you, these are also the questions to ask before hiring anyone to build your systems.
What to look for if you do buy
Whether you can export everything, in a usable form, in one click. This is the only feature that protects you from every other mistake.
Whether it handles the people who are not clients. Many tools built for practices are really scheduling and billing systems, and they are structurally uninterested in the person who inquired in March and did not book. That person is a large part of why you wanted this. So are former clients: most of what the Revenue Engine calls Compound, the repeat work, referrals and reactivation from the people who already know you, depends on being able to find them.
Whether it stores context or only transactions. A record of sessions and payments is an accounting history. It will not tell a colleague anything useful about the person.
And where it lives, and what you would be comfortable telling a client is held in it. A record useful enough to matter is sensitive by definition.
Questions owners ask
What is the best CRM for my kind of business?
There is no answer to that in the abstract, and anyone who gives you one without asking about your business is selling something. Write down the four things above, work out which of them you need software for, then look. Many owners find that two of the four never needed software at all.
Can I just use a spreadsheet?
Yes, for longer than most people expect, and it is what I would suggest starting with. Its limits are real: no automation, poor multi-person editing, and it will not show you patterns across people. When you hit those limits you will know, because they will be causing specific problems you can name.
What about the notes taken during sessions or appointments?
Session notes are a different object from a client record and often belong somewhere separate, particularly in regulated professions with their own requirements for clinical records. The client record holds what the business needs to serve someone well. The session notes hold what happened in the room. Do not collapse them without checking your professional obligations.
Is it worth migrating years of old clients in?
Not all at once, and probably not all of them. Start with everyone active, then work backwards through the people you would genuinely want to speak to again, five minutes each. The exercise of deciding who those are is itself useful, and you will find you remember more than you expect once there is a shape to put it in.